Previous studies documenting associations among violence, substance use, and HIV risk have predominantly relied upon global frequency measures of these events. Whether violence and substance use occur in close temporal proximity with HIV risk behavior has important implications for understanding the determinants of HIV and for the developing effective HIV prevention interventions. The analysis of event-level data can reveal whether potential determinants of HIV risk have a temporally-proximate event-level association with specific instances of risk behavior or have a person-level association with general propensity to engage in risk behavior. The proposed secondary analysis research will evaluate person-level and event-level correlates of sexual risk behavior, substance use, and violence among women at risk for HIV who had recent criminal justice system involvement (n=530). As part of this HIV and violence intervention study (R01DA012572), the Timeline Followback method was used to collect data over 30 days at four different time points on substance use (alcohol use, drug use, injecting, and needle sharing), sexual behavior (type of Intercourse, condom use, partner type, sex exchange, and coerced sex), and experiences of violence (relationship with perpetrator and substance use by victim and perpetrator). Mixed-models will be used to characterize the event-level and person-level associations among substance use, sexual risk behavior, and violence. Additional analyses of critical-incident data will be used to corroborate these findings.